Breaking Down WSU's Resume

[The Mad Hatter]

Maybe. But the primary numbers used to generate the BCS Standings are... you guessed it... the human polls. You wouldn't use that human BS bias here.

The other factor is that basketball offers a much bigger sample size than football with so many more teams and games. This makes it harder to evaluate everyone as an individual, but is better for statistical models that thrive on the additional data points.

[Rlh04d]

Maybe. But the primary numbers used to generate the BCS Standings are... you guessed it... the human polls. You wouldn't use that human BS bias here.

True, but I believe the BCS also uses three computer ranking systems in addition to the two human polls.

I think an issue with that is that they aren't very open about how the computers rank. The basketball rankings I've seen are far more transparent.

[ShockerFever]

The other factor is that basketball offers a much bigger sample size than football with so many more teams and games. This makes it harder to evaluate everyone as an individual, but is better for statistical models that thrive on the additional data points.

Good point.

[Rlh04d]

I have a feeling the Florida issue will work itself out, though, as I see them likely losing to Vandy tonight and Kentucky to close out the regular season, and a mediocre SEC tournament showing.

Florida did better than I thought, but they lose by 10 at Vanderbilt. With Kentucky up next, Florida is going to finish off the season on a three game losing streak.

Still saying that WSU should clearly be a higher seed than UF/FSU.

[rialaigh]

You're right, I did. Forget everything I typed. If Massey says that's where they should be, then I agree with it. Although, other factors do need to be considered... like recent trends, conference records, injuries, suspensions. etc.

I dislike people talking about injuries and suspensions when seeding for the NCAA (however I am fine considering that when ranking). If you take into account individual game preformances and injuries and suspensions and who played and who didn't in the conference tournies than it rewards teams like UNC. Roy Williams will sit most of his starters the first or second game of the ACC Tourny and UNC may go out 2nd round. However the committee will look at that and say...oh, we was just resting his starters, he deserves an even better seed because they are ready to play...and I think rewarding that kind of behavior is bogus. Now I think you can account for that in the rankings but in the seedings I hate seeing that kind of behavior rewarded. It once again is Biased towards BCS teams because if a team like Wichita State were to rest their starters and lose first or second round of the tourny than it would be the end of the world, we would be looking at a 7-9 seed.

[wufan]

My issue is that I don't think the big 10 is as good as the RPI makes them look. My own non-computer bias towards the Big 10 tells me that there are 2-3 really good teams (top 25ish) and 2-3 more that are top 50ish. However, because they play each other so much, it looks like there are all world beaters and gives them 4 top 4 seeds...Not to mention the opportunity to play in the ACC/Big 10 challenge and a pre-season tourney. If you play in a BCS conference, you get an artificially inflated RPI just because you play other Big 10 teams.

So, as a Big 10 team, I could get a middle of the pack ACC team at home and win. That team then gets an inflated RPI because it's in the ACC. Then I play 9 home games against teams in the 150-300 RPI and go 8-1. Play in the Big 10 conference, and I go .500 and everyone maintains their inflated RPI because everyone is showing a winning record with tons of top 50 teams. Basically at 20-10, a Big 10 team is automatically a 4 seed.

No logic, numbers, or examples to back up what I say.:subdued:

[proshox]

Jamar, did you read the shockwaves post regarding road records and seeding?

Since 2007, Butler was the only team with a five seed and only one road loss. They proved deserving of a protected seed. Do you think that you might be under valuing road record compared? BTW, the next worst seed was BYU (3) which had suspended its second best player.

[WuDrWu]

Does the RPI over value a road game and under value a home game? Am I right in thinking it did at some point, or am I thinking about something else?

[DUShock]

I am quite pleased to be considered a four or five seed, regardless of the adjective used.

Go Shocks!

[rrshock]

JH4P you have a heckuva lot more patience and time than me. I know that keeping up with this stuff isn't going to be easy for the next week and a half. I enjoy reading your thoughts on this.

So what does the Florida loss do to it now?

[Jamar Howard 4 President]

Jamar, did you read the shockwaves post regarding road records and seeding?

Since 2007, Butler was the only team with a five seed and only one road loss. They proved deserving of a protected seed. Do you think that you might be under valuing road record compared? BTW, the next worst seed was BYU (3) which had suspended its second best player.

I actually didn't read that article, but I assume I can get the main idea from your post. That's an interesting stat. My initial thought is that neutral games usually get lumped in with road games for a "not at home" type of evaluation. WSU's 1-2 neutral record isn't a strength (not a huge negative either, but atleast a small one). I would be interested to look at all these past one loss road teams and see how they did on neutral courts.

If WSU can win it all in St. Louis, I thnk that helps their chances to get a high seed considerably. At that point, I do think they would get a 4 seed most likely.

[Jamar Howard 4 President]

Does the RPI over value a road game and under value a home game? Am I right in thinking it did at some point, or am I thinking about something else?

Yes, you are correct. The multipliers that they applied to the RPI (1.4 for road wins, 0.6 for home wins, 1.4 for home losses, 0.6 for road losses) are too high. When you look nationally at home records vs away records, the difference is less than what the current RPI formula factors in. As you said, the RPI over values a road game and under values a home game. I think that is part of the reason WSU is overrated (in my opinion) when the RPI has them in the top 10.

[WuDrWu]

#13 Overall - Wisconsin

RPI - 27
Pomeroy - 5
Sagarin - 8
vs 001-025: 3-6
vs 026-050: 3-0
vs 051-100: 2-0
vs 101-200: 9-2
vs 201-345: 4-0
Currently #5 in 1st Ranked Conference
Last 10 games Performance - Very Good
Overall Road/Neutral Performance - Very Good

The RPI doesn't love Wisconsin, but both Pomeroy and Sagarin have them in the top 10. There are very few teams in the country that have only 2 losses outside the RPI top 25, and Wisconsin is one of them. With a solid last 10 games and a solid road/neutral resume (wins @Ohio State & @Purdue), it is hard for me to find many categories where WSU has an edge. Wisconsin could easily lose a couple games and let the Shockers pass them by, but for now, I have Wisconsin in front.

Is 3-6 really that much better than 1-2? I'm not saying Wisconsin is on the outside looking in, but I think too much weight is given to just playing a really good schedule (they're still 3-6 against the RPI top 25, and it's not a stretch at all to say WSU couldn't do at least that good especially if we get an even number of home games). Plus WSU has lost 1/2 the games Wisky has lost including half as many 100-200 games as Wisconsin. That's got to count for something.

And mind you, this is the only selection I disagree with, and it's only by a spot or 2...I just think WSU should be ahead of Wisconsin, at this point. I love the blog JH4P.


And I am still looking for an answer on home and road weighted wins from early. I figure you're the one that knows.

[Jamar Howard 4 President]

JH4P you have a heckuva lot more patience and time than me. I know that keeping up with this stuff isn't going to be easy for the next week and a half. I enjoy reading your thoughts on this.

So what does the Florida loss do to it now?

After looking into Florida State some more, I went ahead and flipped WSU ahead of them. All other changes are due to last night's games. As of today:

3 seeds:
Georgetown, Marquette, Baylor, Michigan

4 seeds:
Indiana, Wisconsin, Louisville, Wichita State

5 seeds:
Florida State, Florida, UNLV, Temple

[Jamar Howard 4 President]

Is 3-6 really that much better than 1-2? I'm not saying Wisconsin is on the outside looking in, but I think too much weight is given to just playing a really good schedule (they're still 3-6 against the RPI top 25, and it's not a stretch at all to say WSU couldn't do at least that good especially if we get an even number of home games). Plus WSU has lost 1/2 the games Wisky has lost including half as many 100-200 games as Wisconsin. That's got to count for something.

And mind you, this is the only selection I disagree with, and it's only by a spot or 2...I just think WSU should be ahead of Wisconsin, at this point. I love the blog JH4P.


And I am still looking for an answer on home and road weighted wins from early. I figure you're the one that knows.

I agree that Wisconsin and WSU are close. Had they not beaten Ohio State on the road this past weekend, I would have WSU in front.

(Look just above your last post for my response about road/home in the RPI.)